Paul Krugman goes all “crowd out” on us. Is he right?

January 9th, 2017 at 5:09 pm

Progressives’ Keynesian economist in chief, Paul Krugman, has been second to none in calling out policymakers’ focus on reducing budget deficits when economies were still weak (also known as “austerity”). Given that record, his oped in today’s NYT may surprise some readers. He argued that, as the economy closes in on full employment, fiscal budget deficits could crowd out private borrowing, pushing up interest rates and slowing growth.

Paul’s argument in the oped shouldn’t actually be surprising; he has long depended on a very simple and, as the record shows, very insightful, application of the ISLM model, a diagram of how interest rates and output interact in key markets in the macroeconomy. See here for his useful discussion of how the model ticks.

But is Paul right? Despite the fact that he invariably turns out to be so–i.e., correct–I’m not nearly so worried about interest-rate crowd-out resulting from the big, wasteful tax cut team  Trump and his Congressional allies will pass, I fear, sometime later this year. What I’m worried out is what their raid on the coffers of the US Treasury will do to the programs we increasingly need to meet the many challenges we face.

Let me explain.

Here at OTE, we maintain that all economic models are wrong but some are sometimes useful. For years, at the end of the ISLM section of economics courses, there’s been this little section that shows how the model changes in a particular type of recession when two things happen: demand significantly contracts and interest rates fall to around zero (the dreaded liquidity trap). At that point you get the diagram Paul put in his link above (ignore for a moment the “IS Now” line, which I plugged in there, as did Paul in a post today).

Source: Paul Krugman

The point of the ISLM-in-recession model is that policy makers can do a lot to boost demand without worrying about crowding out private investment, inflation, or push-back from the Fed. So let the fiscal stimulus rip. The question in such moments shifts from “is the deficit too large?” to “is the deficit large enough?!”

But according to the model, you should only let it rip up until the point when the IS curve shifts enough to the right (“IS Now”) that the economy is back in a place where increased demand will invoke those negative outcomes just noted.

One implication Paul draws from these dynamics is that Republicans, motivated not by improving the economy but by bashing Obama and the D’s, inveighed against deficits when we needed them and are about to shift to not caring about them when deficits – again, according to the model – could actually do some harm.

But how reliable is this crowd-out hypothesis? It’s actually pretty hard to find a correlation between larger budget deficits and higher interest rates in the data.

There are some obvious reasons why that’s the case. Oftentimes, like in the Great Recession, a large budget deficit corresponds with demand contraction and very low rates, so that messes up the predicted correlation. The budget deficit got to -10% of GDP in 2009 and interest rates were stuck around zero. That also implies rates can’t fall as deficits have become less negative.

To see if anything jumps out, the figure plots real rates on the 10-year bond against the deficit from 1990 to 2007, years chosen because the deficit moved around a lot in those years, including into surplus at the end of the 1990s, and the Fed wasn’t nearly as much in the interest-rate setting mix as they’ve been since then. But it’s just pretty much a random plot (if you plot changes in the variables, it still looks random; same with nominal rates; same with corp bonds, etc.).

Source: BEA, Fed

The raw data miss a potentially important expectations component often in play regarding movements in rates. Very recently, investors’ expectations of Trump-induced fiscal expansion, along with the Fed’s plans to hike rates, have pushed up inflation and interest rate expectations. But it’s not at all clear how much of that relates to the expectation that deficits will crowd out private borrowing.

So is Paul making a mistake to continue to depend on the model that has heretofore served him—and anyone else willing to listen—so well? My guess is that deficit crowd-out is not likely to be a big problem, as in posing a measurable threat to growth, anytime soon, even if deficits, which are headed up anyway according to CBO, were to rise more than expected.

The global supply of loanable funds is robust and, in recent years, rising rates have drawn in more capital (pushing out the LM curve). Larger firms have enjoyed many years of profitability without a ton of investment so they could use retained earnings (the fact of unimpressive investment at very low rates presents another challenge to this broad model). And most importantly, while we’re surely closer to full employment, there are still a lot of prime-age workers who could be drawn in to the job market if demand really did accelerate.

(This, by the way, is the only part of Paul’s rap today that I found a bit confusing. He’s a strong advocate of the secular stagnation hypothesis, wherein secular forces suppress demand and hold rates down, even in mature recoveries. His prediction today seems at odds with that view.)

And yet, I’m still really worried—profoundly so—about crowd-out, just not the interest-rate type that Paul’s worried about. What keeps me up at night is that if Republicans are able to waste a bunch of money on deficit-inducing tax cuts that go mostly to rich people, there will be too few resources to support the safety net, public goods, health care, and possibly even social insurance.

This, I’ve long maintained, is the true target of trickle-down tax cuts: force the government to shrink by cutting off its revenue oxygen. And this is a particularly damaging time to be cutting revenues; our demographics alone mean we’re going to need more, not less, revenues in coming years. And I’m not even talking about what we’ll need to address the challenges posed by climate change, inequality, poverty, our infrastructure needs, geopolitics, and Buddha-knows what else.

So I stand firmly against big, dumb wasteful tax cuts. Not because I think they’re going to raise interest rates that much (though I could be wrong and PK is typically right), but because they’re going to shut down the federal government’s ability to do what needs to be done.

Print Friendly, PDF & Email

18 comments in reply to "Paul Krugman goes all “crowd out” on us. Is he right?"

  1. Gerald Scorse says:

    “What keeps me up at night is that if Republicans are able to waste a bunch of money on deficit-inducing tax cuts that go mostly to rich people, there will be too few resources to support the safety net, public goods, health care, and possibly even social insurance.” This is absolutely the right thing to worry about. Crowd-out? If it even happens, it’s small potatoes next to the paring down of the safety net etc.

    • Andrew says:

      No, no, no. Does Bernstein STILL not understand that the US is monetarily sovereign and can provide tax cuts and services if it so chooses? What is the limiting factor? Just the will of congress. Now THERE is our big problem.

  2. Bill says:

    No he’s a mile off

    Just look at the graphs

    Somebody needs to tell Paul the FED sets the rate.

  3. Smith says:

    We are not at full employment. We are not close to full employment. It would take five more years of 180,000 per month new jobs without interruption to be near full employment. That’s 5 * 12 months * 80,000 above natural population growth, = 4.8 million. That will partly cover the 2% drop in prime age labor force participation fall, (2.5 million if you look at narrow 25 to 54 year old range), and 1.5% extra underemployed (of 150 million workforce, that’s 2.25). Is my math correct? I think.
    Let’s pretend Krugman isn’t stupid. Let’s hope he’s spreading fake economic news about the unemployment situation to give cover to those trying to defeat Trump’s likely successful effort to stimulate the economy. In November he still said Clinton’s infrastructure plan should be larger.
    Likewise wages and quits. Others have seen this too and provided links. Look at these:
    Check out Chart 5. Quits and hires are at pre-recession levels. As if there was no growth in the economy since then. Job openings are higher but not for hire, keeping perpetual openings unfilled is the new normal, another way to control workers, always having replacements at the ready. Keeps HR people employed and busy too. I hear HR people have a lot of say in the hiring process. Funny that they would make more work for themselves, keep themselves in demand.
    We remind our readers too of the strange excitement over wages, going up 1%, skewed to the higher income earners, you reported the bottom 80% actually only got . 5% real raise. Need I remind you the State of Working America lamented stagnant wages for college degree wage earners in the 2000s by showing the most successful segment, Tech workers, managed to average only .3% up to the recession. Again, core inflation is actually above 2%, inflation held in check by energy, and food and low interest rates up to now.

    Pretty pathetic when even liberals shake their heads and say this is the best we can do. What about a law to stop profitable factories from closing or moving? What about ending the use of unemployment (interest rates) to control business raising prices (inflation0? What about ending exploitation of foreign labor? What about breaking up the to big to fail banks?

    • urban legend says:

      Full employment?

      Between 2nd Q 2000 and 4th Q 2016:

      Working age population up 20%
      Part-time jobs up 23%
      Full-time jobs up 8.8%

      And this doesn’t even address what percentage increase in quality full-time jobs (with benefits) we would see — probably a lot less than 8.8%.

  4. Prairie Populist says:

    A couple of observations:

    We may be approaching full employment, but full employment should be celebrated, not feared. What concerns us about full employment is that wage-push inflation will soon follow, then wage price spirals, then gas lines and WIN buttons and all of that. But I think that will not happen this time. There are few COLAs in union contracts, heck, there are few union contacts at all. Labor has no bargaining power. Without that, there will be no wage-price spiral.

    Krugman argues that Republicans will quite possibly do something reckless and drive up interest rates. But stomping on the brakes now, to mitigate the harm the\ Republicans might do, will hurt ordinary citizens. Better to wait and see what happens.

    I share your concern that “starve the beast” is behind a lot of the Republican fiscal irresponsibility. We are in the middle of a very serious and expensive medical crisis here. Thanks to Medicare and an Advantage Plan, we can afford it. If Ryan voucherizes Medicare, we’re bankrupt.

    • James B says:

      The unspoken horror of this inflation concern is that this concern itself causes unemployment, and this is never spoken about in polite company.

      The received wisdom (from, as you say, a very different era) is that if unemployment gets too low then we should raise rates, but what is not said loud enough is that the specific purpose of the rate increase is to re-unemploy some other people in order to ‘cool things down’.

      Thus the use of rates in this way has the explicit purpose of deliberately unemploying people who would otherwise be employed. The trigger is too many people working, and the trigger will be released when rates have had their effect and fewer people are working again.

      Thus we deliberately consign millions of people to the impossibility of finding a job, and then the rest of society berate them as being lazy and not ‘deserving’ of any social support.

      There has to be a better way.

  5. Arun says:

    To see the effect of the annual deficit, shouldn’t you use a shorter term of maturity of bonds, rather than 10 years? Don’t typical corporate projects want Return on Investment within an year or two? Why would they look at the 10 year rate to decide on whether to invest (which is the crowding out effect that you want to find)?

    I tried doing it myself but navigating the data on the Federal Reserve site is difficult for me.

    • Arun says:

      I am no economist, but the more I think about it, using the 10-year rate as the response to current federal deficits seems more and more flaky.

      The 10-year rate is some kind of amalgam of expectations across 5 Congresses and at least 2 different Presidents, if it were possible to disentangle the effect of one year’s deficit on the 10-year rate, it would be truly remarkable.

      I’m no economist, but to convince me you have a case, you need to show me that there is no signal in a measure that has the potential to show a signal. The deficit/GDP vs the real 10 year leaves me very skeptical.

  6. Flaky lefthander says:

    While I appreciate your point that after tax cuts the right side will try to cut critical services, what needs to be said is that you cannot kill the federal government’s ability to do what needs to be done. If you explain that, people will listen to you. And please don’t call PK progressive. That’s fake news.

  7. SteveD says:

    1. The U.S. federal government doesn’t borrow, nor does it need to. What’s incorrectly termed “federal borrowing” is nothing more than the issuance of T-securities. Federal “debt” is the total of T-security accounts at the Federal Reserve Bank — similar to bank savings accounts.
    The “crowding out” hypothesis proposes that these FRB savings account deposits crowd out lending, actually an IMPOSSIBLE notion.
    2. The federal government pays ALL its bills by creating new dollars, ad hoc. To pay a creditor, each federal agency sends instructions (not dollars) to the creditor’s bank, instructing the bank to increase the balance in the creditor’s checking account. When the bank does as instructed, dollars are created and added to the economy.
    3. Interest rates are determined by the Fed, and evolve from the federal funds target rate, which the Fed arbitrarily sets.
    4. Issuing T-securities cannot “absorb the economy’s lending capacity.” The federal government DOES NOT BORROW FROM BANKS. The federal government issues T-securities, having no effect on any bank’s lending capacity (which is determined solely by bank capital).

    • JF says:

      Well this is just nutty. Please source every statement you have made here.

      The Treasury places their borrowing actions into a public auction. Should a bank or any other entity including any natural person wish to purchase, except the Fed Reserve, that person or entity must have collected the monies (somehow) and transfer these to the Treasury, who provides them with what is referred to as a Treasury security (bond, bill, other instrument). The Fed is not permitted by law to purchase Treasury debt issuances directly.

      The Treasury causes lawful claims against the govt to be paid by check or electronic funds transfers of money, just like any other person. The incoming funds come from borrowing and from all ways and means. The Treasury does not pay claims by way of printing new money. The public’s government extracts a flow of monies from the economy and uses it to pay on lawful claims. The Treasury, again, does not print money to pay claims.

      Please read real sources, whatever sources you are using now are completely misleading you.

      I only saw one point in your comment that is accurate, I wonder if you know which one?

  8. Delboy says:

    Bill Mitchell has done a post today that destroys Krugman

    And explains in detail why Krugman has it so wrong.

    It’s superb !

    • Smith says:

      Interesting read, I was not familiar with Bill Mitchell previously. He also thinks we are not near full employment and gives this link in the link you gave. I have run across Robert Farmer who argues against the natural rate of unemployment hypothesis with implications for not raising interest rates.
      And then of course there is me, arguing too, against statements that allege we are nearing full employment, statements from Krugman, and this blog too.
      I give data, I cite sources, I google extensively to provide them, and check my math.
      Ironically, there is much I picked up from Krugman’s blog and this blog, which inform my analysis.
      Trump had no trouble recognizing unemployment figures didn’t accurately reflect the state of the economy.
      When Trump is right and Krugman is wrong, you know you’re in trouble. It’s doubtful Krugman knows many or any people stuck in U6 limbo.

    • rjs says:

      here’s krugman, 5 months ago, explaining why krugman is wrong today:

  9. PJR says:

    Seems to me that if the government reduces taxes by $X billion on the investor class, but borrows $X billion more from the investor class, “crowding out” and interest rate pressures are not major concerns. The government would, however, have to pay interest on the borrowed money, reducing the amount available for other purposes.

    • JF says:

      Yes, thank goodness. Common sense. No finances crowded, and by definition if you chose instead to invest in some other asset than a govt bond, then clearly you are not crowded out.

      But also don’t forget that the bond traded for cash is still a highly liquid asset that now has incredible use as a medium of exchange in the financial asset trading market places. You are even better enabled to trade in these marketplaces, no crowding here either.

      So yes, the tax contributions you were going to make before the law was changed in your favor (trading cash to fulfill the tax bill) ends up being exchanged for a security or wealth-asset that also grants you control over future income too. Such a deal, two for one or better even. How do I get in that line?

      I just hope common sense lights up, so this practice is halted.

  10. Dave says:

    Is Krugman trying to appeal to VSPs?

    I agree with you here. But what question are we trying to answer?

    Is trickle down economics going to work again? No, it never worked. Does the size of government have an intrinsic effect upon the economy? No.

    So we need the government we need. The size is not important if it is democratically controlled.

    Banks love to lend money. They will lend to people or government. They don’t discriminate. Are there enough lendable funds? Oh yes, there’s an almost infinite well of money regardless of interest rates.

    Good post, Jared.